Happy moments are full of our daily lives, and today we have a chance to dive into a handy dataset that can help shed some light on the fundamentals of happiness. As for me, I am wondering people under different ages would have differnet happy moments or not. So I decide to analyze the happy moments of different age groups. I separate the data into four age groups.
First group is people whose age is less an 25 years old. They represent youth such as students.
Second group is people who are between 25 to 40 years old.They represent early adulthood.
Third group is people with age between 40 to 60 years old. They represent middle adulthood.
Fourth group is people whose age is above 60 years old. They represent the elderly, late adulthood.
In order to get some basic characteristics of happy moments from these four groups, the word count, word cloud, and bigram network would be analyzed.
Youth
Early Adulthood
Middle Adulthood
Late Adulthood
Youth
Early Adulthood
Middle Adulthood
Late Adulthood
So far we’ve considered words as individual units; However, many interesting text analyses are based on the relationships between words. Here I arrange the word groups into a network, as a combination of connected nodes.
It has three variables:
From: the node an edge is coming from
To: the node an edge is going towards
Weight: A numeric value associated with each edge
Youth
Early Adulthood
Middle Adulthood
Late Adulthood
We have some interesting moments happened. People usually focus on different things when they under different age periods. As for youth who study at school, they think friends, family, mother’s day, watch TV, play games, go to events, have birthday party, hang out with girlfriend/boyfriend or go to school/college can make them happy.
When people get into early adulthood, they may start getting a job, and many of them get married. They think friends, marriage, family(mother’s day, son/daughter, husband/wife), have a birthday party, watch TV, play video game can make them happy.
When people enter middle age, they may have a job for a long time and they feel that family(son/daughter, wife/husband), friends, celebrate birthday party, watch TV, read books, or cook dinner can make them happy.
After they retire, friends and family(son/daughter, wife/husband, grandson/granddaughter) are most important for them, they feel happy when they walk with dog, watch movie, visit or call someone.
All in all, we find out that spend time with family/friends, to have mother’s day, birthday party, and play/watch games are the most happy moments ever.
In this part, I apply word frequences to analyze, and using a correlation test to quantify how similar and different these sets of word frequencies, how correlated are the word frequencies.
Correlation test
##
## Pearson's product-moment correlation
##
## data: proportion and Above 60
## t = 62.628, df = 2375, p-value < 2.2e-16
## alternative hypothesis: true correlation is not equal to 0
## 95 percent confidence interval:
## 0.7735516 0.8039080
## sample estimates:
## cor
## 0.7892114
Correlation test
##
## Pearson's product-moment correlation
##
## data: proportion and Above 60
## t = 100.5, df = 2626, p-value < 2.2e-16
## alternative hypothesis: true correlation is not equal to 0
## 95 percent confidence interval:
## 0.8826954 0.8984946
## sample estimates:
## cor
## 0.8908641
##
## Pearson's product-moment correlation
##
## data: proportion and Above 60
## t = 125.71, df = 2338, p-value < 2.2e-16
## alternative hypothesis: true correlation is not equal to 0
## 95 percent confidence interval:
## 0.9279151 0.9383742
## sample estimates:
## cor
## 0.9333424
Correlation test
##
## Pearson's product-moment correlation
##
## data: proportion and Below 25
## t = 268.6, df = 6545, p-value < 2.2e-16
## alternative hypothesis: true correlation is not equal to 0
## 95 percent confidence interval:
## 0.9554471 0.9594789
## sample estimates:
## cor
## 0.9575098
##
## Pearson's product-moment correlation
##
## data: proportion and Below 25
## t = 114.66, df = 4536, p-value < 2.2e-16
## alternative hypothesis: true correlation is not equal to 0
## 95 percent confidence interval:
## 0.8545936 0.8695309
## sample estimates:
## cor
## 0.8622496
Correlation test
##
## Pearson's product-moment correlation
##
## data: proportion and Between 40 to 60
## t = 239.45, df = 5395, p-value < 2.2e-16
## alternative hypothesis: true correlation is not equal to 0
## 95 percent confidence interval:
## 0.9536778 0.9582699
## sample estimates:
## cor
## 0.9560324
We make a camparison between different age groups to figure out wheater they have same happy moments or not.
First round, I am interested in comparing people under 25 years old with people whose age is above 60 years old.They have the greatest age difference, and have the lowest correlation. From the analysis, we can see that they do not have many common happy moments besides the moment with family or friends. Somewhat surprisingly, walk dog, play golf, garden/yard, neighbor, daughter,granddaughter can make the elderly happy, but young people(students) think that play game, college/class, and girlfriend/boyfriend make them happy.
Second round, both of them feel that family(such as daughter), call someone, or buy something can make them happy. Only one thing is different, grandchild appers in the elderly’s (age above 60 years old) happy moments. Also, we realize that late adulthood has more higher correlation with middle adulthood than with early adulthood.
Third round, they both think tht have a birthday party makes them happy. But early&middle adulthood also feel that son/daughter or even grandchildren make them happy, this is the moment youth do not have.
Fourth round, early adulthood and middle adulthood may both enter the workforce and have many similar happy moments. Thus, they have the highest correlation.
Overall, during different age periods, people have different happy moments. If age difference is smaller, they have more similar happy moments because the happy moments’s correlation is higher; However, two moments are important for every age stage: FRIEND and FAMILY/HOME; They make people happy no matter who you are, and no matter how old you are.
Sentiment analysis weighs the emotional intensity of text and I measure the sentiment of happy moments to see how their intensities vary from different age stages.
## Min. 1st Qu. Median Mean 3rd Qu. Max.
## -3.2000 0.2500 0.8000 0.9878 1.5000 22.4000
## Min. 1st Qu. Median Mean 3rd Qu. Max.
## -5.4000 0.2500 0.8000 0.9872 1.5000 22.7000
## Min. 1st Qu. Median Mean 3rd Qu. Max.
## -3.8000 0.2500 0.8000 0.9695 1.5000 17.1500
## Min. 1st Qu. Median Mean 3rd Qu. Max.
## -2.500 0.300 0.800 1.003 1.550 20.550
Before doing this sentiment analysis, I am confident that they all should be positive since it is the data about happy moments. However, the result surprised me.
Exactly, overall the moments are definitely positive and the 25th, 50th, and 75th percentiles across age groups are virtually identical. But the bottom quartile does have negative sentiment. I guess the reason is that some happy moments may poetically arise from discomfort things.
Hi, here.
Our journey of exploration is coming to the end. I feel so glad to realize that no matter how old you are, happy moments are always around us. Maybe you will change some hobbies when you get older, maybe you will have more family mombers,more friends. But one thing is elernal, that is the love from your family and friends. I trust that they can give you the most happiest moments forever.
Thank you for reading ^^